An Adaptive Detection Method of Spatial Circular Feature Based on Arc Segment Under Different Lighting Conditions

Accurate recognition of spacecrafts is an important prerequisite for on-orbit service (OOS). For missions such as on-orbit operation, rendezvous and docking, structures with special shape such as docking ring or engine nozzle are usually selected as recognition objects, whose circular feature can help to accurately locate and determine the potential operating points for subsequent tasks. Different lighting conditions in space are easy to interface with visible light imaging, resulting in different brightness, loss of details and blurring of some key features. This paper proposed an adaptive detection method of spatial circular feature based on arc segment to solve the problems of accurate recognition of spacecraft’s docking ring under low illumination and local over-exposure conditions. Firstly, we designed an adaptive image enhancement pre-processing method based on multi-scale Retinex with chromaticity preservation (MSRCP). Images obtained by visible camera under different lighting conditions were enhanced by MSRCP and then processed by adaptive threshold method for segmentation, filtering and edge detection. Then an ellipse detection algorithm based on arc segment features was designed to extract the contour of the target's docking ring from the edge curve. Finally, a ground simulation experimental system was built, and we carried out some experiments under different lighting conditions. The experimental results showed that our method can effectively improve the image quality under low illumination and local over-exposure conditions. The average brightness, standard deviation and information entropy of images in different conditions were improved. The circular feature detection algorithm showed excellent detection results on the enhanced images under different lighting conditions. This paper provides a certain reference for the accurate recognition of targets in on-orbit service missions.